( 864 ) 
possibility of liquefying helium an open question at that time. 
To test the apparatus, expansion experiments were first to be made 
with air at ordinary temperature, and with hydrogen at the temperature 
of liquid oxygen; but in the construction of the apparatus it had 
always to be kept in view that it was eventually to be used in 
liquid hydrogen. Preliminary experiments showed that the apparatus 
in its original form did not give a true JouLE-KeLvin-effect, and thus 
led to a special investigation of the thermodynamics of expansion 
through a valve and how a valve apparatus has to be arranged so 
that the enthalpy‘) before and after expansion remains the same. 
The results of this investigation will be published in this and following 
communications, for they lose none of their interest by the circum- 
stance that the original object of the research has been disposed of 
by the liquefaction of helium. *) 
$ 2. Since the experiments of Joure and Ketvin *) of 50 years 
ago very little experimental work upon similar expansions of gases has 
been published. This lack of confirmation is rather surprising when 
we consider the importance of the JouLE-KrLVIN experiments in gas- 
thermometry and the frequency with which their results are employed 
in various theoretical thermodynamical investigations. With the 
exception of some rough experiments by REGNAULT *) upon various 
gases, and the work of EK. Natanson *) and Kester *) upon CO,, no 
further measurements of the JouLz-Kenvin effect seem to have been 
made. Reanavtt did not obtain results sufficiently definite to lead to 
any theoretical conclusions, and the results obtained by the other 
experimenters are not in agreement. 
§ 3. Although Joure and Kervin began their experiments by 
allowing the expansion to take place through a small aperture, they 
soon abandoned that form of apparatus, and used instead a porous 
plug so as to ensure, by friction in the plug, the immediate conver- 
1) This name has been suggested by KamerrineH Onnes to indicate the function 
(e+ pv) — the “Heat function” of Graas. H. L. Carrenpar (Phil. Mag. [6]. 5. 
p. 48. (1903).) calls this expansion *Adiathermal”’. 
2) H. KAMERLINGH ONNes: Comm. Phys. Lab. Leiden. N’. 108. (Proc. June 1908). 
3) Joure and Kervin: Phil. Mag. [4]. 4. p. 481. (1852); Phil. Trans 143. p. 357. 
(1853); 144. p. 321. (1854); 152. p. 579. (1862). 
+) V. Reenautt: C. R. 69. p. 780 (1869). 
5) E. Naranson: Wied. Ann. 31 p. 502 (1887). 
6) F. KE. Kester: Physik. Zeits. 6 p. 44 (1905). 
Phys. Rev. 21 p. 260 (1905). 
